We have used DNA microarrays to survey rates of mRNA decay on a genomic scale in early stationary-phase cultures of Bacillus subtilis. The decay rates for mRNAs corresponding to about 1500 genes could be estimated. About 80% of these mRNAs had a half-life of less than 7 min. More than 30 mRNAs, including both mono- and polycistronic transcripts, were found to be extremely stable, i.e. to have a half-life of > or =15 min. Only two such transcripts were known previously in B. subtilis. The results provide the first overview of mRNA decay rates in a gram-positive bacterium and help to identify polycistronic operons. We could find no obvious correlation between the stability of an mRNA and the function of the encoded protein. We have also not found any general features in the 5' regions of mRNAs that distinguish stable from unstable transcripts. The identified set of extremely stable mRNAs may be useful in the construction of stable recombinant genes for the overproduction of biomolecules in Bacillus species.

The Bacillus subtilis aprE leader is a determinant of extreme mRNA stability.The authors examined what properties of the aprE leader confer stability on an mRNA. The secondary structure of the aprE leader mRNA was analysed in vitro and in vivo, and mutations were introduced into different domains of an aprE leader-lacZ fusion. The half-lives of the corresponding transcripts were determined and β-galactosidase activities were measured. Removal of a stem-loop structure at the 5' end or diminishing the strength of the RBS reduced the half-lives from more than 25 min to about 5 min. Interfering with translation by abolishing the start codon or creating an early stop codon had no or little effect on mRNA stability. The authors conclude that a 5' stem-loop and binding of ribosomes are necessary for the stability of aprE leader mRNA. The present results, together with a number of other data, suggest that translation of a B. subtilis mRNA is generally not important for its stability ; the situation seems different in Escherichia coli. It is further concluded that the calculated strength of a B. subtilis RBS cannot be used to predict the stability of the corresponding transcript.

The Bacillus subtilis aprE gene encodes subtilisin, an extracellular proteolytic enzyme produced in stationary phase. The authors examined the stability of aprE mRNA and aprE leader-lacZ fusion mRNA. Both mRNAs were found to be unusually stable, with half-lives longer than 25 min, demonstrating that the aprE leader contains a determinant for extreme mRNA stability. The half-lives were the same in growing and stationary-phase cells. This contrasts with the findings of O. Resnekov et al. (1990) [Proc Natl Acad Sci U S A 87, 8355-8359], which suggested a growth-phase-dependent mechanism for decay of aprE mRNA. The discrepancy is explained by the techniques used. Substitution of two bases or deletion of 25 nucleotides in the aprE leader led to a major difference in its predicted secondary structure and resulted in a fivefold reduction of the half-life of aprE mRNA. The authors also determined the halflife of amyE mRNA, which encodes α-amylase, another stationary-phase, excreted enzyme and found it to be around 5 min. This shows that extreme stability is not a general property of stationary-phase mRNAs encoding excreted enzymes.

The green fluorescent protein GFPuv has been genetically engineered to investigate the influence of N-terminal tyrosine extensions in aqueous two-phase systems. Fusions in the N-terminus affected the protein expression, and tags containing three tyrosines and prolines influenced the expression favorably. This effect is probably due to changes in mRNA stability, because the amounts of corresponding mRNAs correlated with the amounts of GFPuv proteins. The partitioning was investigated in two different aqueous two-phase systems, a two-polymer system composed of EO30PO70/dextran and a PEG/salt system with potassium phosphate. Partitioning in the PEG/salt system generally was more favorable than in the EO30PO70/dextran system. Tags with three tyrosines resulted in higher partitioning toward the EO30PO70- and PEG-rich phases, respectively. The effect of adding proline residues to the tag was also investigated, and the partitioning effect of the tag was enhanced when prolines were included in the tags with three tyrosines. The best tyrosine tag, Y3P2, increased the partition coefficient 5 times in the PEG/salt system. Thermoseparation of the EO30PO70 phase allowed recovery of 83% Y3P2-GFPuv protein in a water phase.

Conversion of undesirable, taste-active compounds is crucial for using barley as a suitable raw material for beer production. Here, ALH1, a barley alkenal hydrogenase enzyme that reduced the alpha,beta-unsaturated double bond of aldehydes and ketones, was found to convert trans-2-nonenal (T2N), a major contributor to the cardboard-like flavor of aged beer. Although the physiological function of ALH1 in barley development remains elusive, it exhibited high specificity with NADPH as a cofactor in the conversion of several oxylipins-including T2N, trans-2-hexenal, traumatin, and 1-octen-3-one. ALH1 action represents a previously unknown mechanism for T2N conversion in barley. Additional experimental results resolved the genomic sequence for barley ALH1, as well as the identification of a paralog gene encoding ALH2. Interestingly, T2N was not converted by purified, recombinant ALH2. The possibility to enhance ALH1 activity in planta is discussed--not only with respect to the physiological consequences thereof--but also in relation to improved beer quality.